Reduction of noise effects in imperfect graph states

Google meet: https://meet.google.com/upw-ynon-pkx

Abstract

Graph states are regarded as a testing ground of various quantum information schemes – they are useful in analysis of e.g. cryptography, measurement-based quantum computation, error correction, and metrology. As a result of their simple entanglement structure, multiple physical implementation exist, including ion traps, photonic qubits, and superconducting qubits. However, such realizations invariably suffer from noise, which presents itself in a multitude of ways: the engineered interactions may have imperfect strengths, additional transitions may arise, and the systems rarely can be protected from coupling with the outside world. All of these contribute to the reduced utility of the produced quantum states. Despite different physical origins, various noise processes often lead to a similar structure of the resulting state: it is a mixture of stabilizer states. This observation can be exploited in order to detect – and reduce – the effects of noise, without the large overhead of full error correcting code. I exemplify this approach with a prototypical task of extraction of a Bell pair across distant parts of the graph. Entanglement quality of the resulting state, directly measuring its utility for further information processing, can be maintained in the presence of noise even for large distances for some geometries of the underlying graph.